In the escalating debate over artificial intelligence infrastructure’s environmental impact, “Shark Tank” investor Kevin O’Leary is deploying a counterintuitive argument: AI data centers actually consume less water than America’s golf courses. But while technically accurate on surface level, the comparison obscures a far more complex picture that tech industry critics say glosses over legitimate sustainability concerns.
What Happened
O’Leary’s 40,000-acre Stratos data center project in Utah has faced significant opposition, including protests and an executive order from Utah’s governor. In response to mounting environmental scrutiny, O’Leary told Business Insider that AI facilities use substantially less water annually than the nation’s approximately 15,000 golf courses. The claim, while numerically defensible, has reignited fierce debate about how the industry frames its environmental footprint.
Key Points
The numbers tell a selective story. U.S. golf courses collectively use roughly 2 trillion gallons annually, far exceeding what most individual data centers consume. However, critics argue this comparison is fundamentally misleading. Data centers are concentrated industrial facilities in water-stressed regions, while golf courses, though water-intensive, distribute their usage across the country. A single massive data center in Utah—an arid state already battling drought conditions—poses different challenges than dispersed golf course irrigation.
Water usage also represents just one environmental metric. Data centers demand enormous electrical power, cooling infrastructure, and land development. The comparison conveniently sidesteps these broader sustainability questions while choosing a convenient industry punching bag in golf courses.
Tech companies increasingly tout similar arguments to deflect environmental concerns. Apple, Meta, and Microsoft have made comparable claims, suggesting their operations pale compared to agriculture or other industries. While not dishonest, such comparisons strategically redirect conversations away from the genuine resource demands of AI infrastructure expansion.
What This Means
For the tech industry and investors pushing AI expansion, these framing tactics matter enormously. As data center projects proliferate across water-vulnerable regions, companies will likely escalate comparative arguments to neutralize local opposition and regulatory scrutiny. O’Leary’s approach represents a calculated PR strategy: acknowledge environmental concerns while presenting an alternative metric that makes AI infrastructure appear benign.
For policymakers and communities, the lesson is clear: single-metric comparisons obscure complexity. Evaluating data center projects requires comprehensive environmental assessments addressing water depletion, energy demand, grid stability, and cumulative regional impact—not selective comparisons designed to mollify opposition.
As AI infrastructure investment accelerates, expect more such arguments. The real question isn’t whether data centers use less water than golf courses. It’s whether they’re sustainable given their concentrated environmental footprint in increasingly water-stressed regions.
